HOPE VI Data Compilation and Analysis
PD&R Research Partnerships PD&R Research Partnerships Partnering With Universities, Philanthropic Organizations, Other Federal or State Agencies for Innovative Research Projects That Inform HUD’s Policies and Programs
HUD’s Office of Policy Development and Research (PD&R) has authority to enter into unsolicited research partnerships
with universities, philanthropic organizations, other federal or state agencies, or a combination of these entities through
noncompetitive cooperative agreements. The purpose of these partnerships is to allow PD&R to participate in innovative
research projects that inform HUD’s policies and programs. Research partnerships require that at least 50 percent of the
costs are funded by the partnering agency. PD&R is focusing its cooperative agreement efforts on research partnerships
that will advance one of the following five key areas:
1. Homeownership and housing finance;
2. Affordable rental housing;
3. Housing as a platform for improving quality of life;
4. Sustainable and inclusive communities; or
5. HUD assets.
Visit PD&R’s website www.huduser.gov to find this report and others sponsored by HUD’s Office of Policy Development and Research (PD&R). Other services of HUD USER, PD&R’s research information service, include listservs, special interest reports, bimonthly publications (best practices, significant studies from other sources), access to public use databases, and a hotline (800-245-2691) for help accessing the information you need.
Final Report submitted to: Office of Policy Development and Research U.S. Department of Housing and Urban Development
Cover photos courtesy of McCormack Baron Salazar featuring HOPE VI projects Murphy Park in St. Louis, MO, former C.J. Peete, now Harmony Oaks, in New Orleans, LA and Tremont Pointe in Cleveland, OH.
This report was supported by funding under a Research Partnerships grant (RP-14-OH-005) with the U.S. Department of Housing and Urban Development. The substance and findings of the work are dedicated to the public. The author and publisher are solely responsible for the accuracy of the statements and interpretations contained in this publication. Such interpretations do not necessarily reflect the views of the Government.
National Initiative on Mixed-Income Communities 2 | P a g e Case Western Reserve University Table of Contents List of Tables ...... 5 List of Figures ...... 6 I. Introduction ...... 7 Policy Context ...... 7 Knowledge Gap ...... 7 National Initiative on Mixed-Income Communities...... 7 Project Description ...... 8 II. Summary of Key Findings ...... 11 Redevelopment Progress ...... 11 Unit Production ...... 11 Return and Relocation ...... 12 Timeframes of Redevelopment Progress ...... 12 Financing ...... 13 Community and Supportive Services ...... 13 III. Findings ...... 15 Grant Awards ...... 15 Nationwide ...... 15 HOPE VI Grants by Region ...... 18 HOPE VI Grants by State ...... 20 Redevelopment Progress ...... 24 Redevelopment Progress by Award Year ...... 24 Income/Tenure Mix ...... 26 Unit Production by Subsidy Type ...... 29 Tenure Mix ...... 29 Projected versus Actual Production ...... 30 Timeframes of Redevelopment Progress ...... 32 Comparative Analysis of Redevelopment Progress ...... 34 Region ...... 34 Income/Tenure Mix ...... 35 Tenure Mix ...... 37 Size ...... 38
National Initiative on Mixed-Income Communities 3 | P a g e Case Western Reserve University Age ...... 39 Return Rates...... 40 Return Rates by Award Year ...... 40 Comparative Analysis of Return Rates ...... 42 Region ...... 42 Income/Tenure Mix ...... 42 Size ...... 43 Age ...... 44 Funding Sources...... 44 Comparative Analysis of Funding ...... 47 Region ...... 47 Income/Tenure Mix ...... 48 Age ...... 48 Community and Supportive Services (CSS) ...... 50 Nationwide ...... 52 CSS Services across Sites ...... 55 IV. Discussion ...... 56 Production...... 56 Financing ...... 57 Community and Supportive Services ...... 58 Study Limitations ...... 58 V. Conclusion and Implications for Research and Policy ...... 59 Implications for Policy ...... 59 Implications for Further Research ...... 60 Acknowledgments ...... 62 References ...... 63
National Initiative on Mixed-Income Communities 4 | P a g e Case Western Reserve University List of Tables Table 1: Comparison of projected and actual unit production ...... 17 Table 2: Relocation, evictions, occupancy and re-occupancy ...... 17 Table 3: Financial information by HUD regions ...... 19 Table 4: Unit construction information by HUD regions...... 20 Table 5: Implementation grant information by states ...... 21 Table 6: Unit production information by states ...... 22 Table 7: Relocation timeframes ...... 33 Table 8: Demolition timeframes ...... 33 Table 9: Construction timeframes ...... 33 Table 10: Occupancy timeframes ...... 33 Table 11: Percentage of unit production by region ...... 34 Table 12: Comparison of average unit production by income mix ...... 36 Table 13: Comparison of average percentage of actual unit production by income mix ...... 36 Table 14: Comparison of average unit production by tenure mix ...... 37 Table 15: Comparison of average percentage of actual unit production by tenure mix ...... 37 Table 16: Average percentage deficit of production by tenure mix ...... 38 Table 17: Comparison of average unit production by size ...... 38 Table 18: Comparison of average percentage of actual unit production by size...... 38 Table 19: Comparison of average unit production by age ...... 39 Table 20: Comparison of average percentage of unit production by age ...... 40 Table 21: Comparison of average rates by region ...... 42 Table 22: Comparison of return rates by income mix ...... 43 Table 23: Comparison of return rates by tenure mix ...... 43 Table 24: Comparison of return rates by size ...... 43 Table 25: Comparison of return rates by age ...... 44 Table 26: Funding Sources across Developments ...... 44 Table 27: Comparison of average HOPE VI funding and total funding by region ...... 47 Table 28: Comparison of average HOPE VI and total funding by income mix ...... 48 Table 29: Comparison of average HOPE VI and total funding by tenure mix ...... 48 Table 30: Comparison of average HOPE VI and total funding by age ...... 49 Table 31: Nationwide CSS information ...... 52 Table 32: Nationwide CSS caseload of pre-revitalization residents and new residents...... 54 Table 33: CSS Services ...... 55
National Initiative on Mixed-Income Communities 5 | P a g e Case Western Reserve University List of Figures Figure 1: Number of grants by award year ...... 16 Figure 2: HUD’s Regions ...... 18 Figure 3: Redevelopment progress by award year ...... 24 Figure 4: Cumulative redevelopment progress by award year ...... 25 Figure 5: Cumulative demolition and three types of construction by award year ...... 26 Figure 6: Income mix across 259 HOPE VI developments ...... 27 Figure 7: Tenure mix across 259 HOPE VI developments ...... 28 Figure 8: Rental and homeownership unit production by award year ...... 30 Figure 9: Count and proportion of mix type across developments...... 35 Figure 10: Return rates by award year ...... 41 Figure 11: Expended HOPE VI funding by award year ...... 45 Figure 12: Expended funding sources by award year ...... 46
National Initiative on Mixed-Income Communities 6 | P a g e Case Western Reserve University I. Introduction
Policy Context As HUD advances its mission to create strong, sustainable, inclusive communities and quality, affordable homes for all, the promotion of mixed-income communities has become a core strategy. Across the U.S., local governments and private developers are increasingly turning to mixed-income development as an approach to deconcentrate poverty and revitalize urban neighborhoods.1 With the Choice Neighborhoods Initiative, launched in 2010, the federal government has extended its commitment to supporting the mixed-income approach to public housing transformation that was first implemented through the HOPE VI initiative in the mid- 1990s. With the dramatic decrease in public sector funding for public housing, housing authorities are increasingly turning to the privatization of public housing through mixed-income development, and more recently through the Rental Assistance Demonstration (RAD) program, as a means of generating the capital needed to construct new buildings and renovate existing ones, as well as providing the operating capital to manage and sustain them.
Knowledge Gap Despite an extensive literature of evaluation reports and articles on various aspects of the $6 billion HOPE VI program2, there has been no detailed descriptive analysis of the overall production of mixed-income units through this effort. Most of the research and analysis on the HOPE VI program consists of studies of a single HOPE VI site or a selected subset of sites. The best available information on the full HOPE VI grant portfolio was generated by Tom Kingsley of the Urban Institute as a part of a comprehensive volume on HOPE VI edited by former HUD Secretary Henry Cisneros and Lora Engdahl, which was published in 2009. In that volume, data are provided on 240 revitalization grants from 1993-2007. However, while total overall counts are provided of units demolished, constructed and occupied by income subsidy level, the development-by-development data rely on projections of expected units, not information about completed units. Furthermore, there have been at least seven additional years of unit production since that analysis was completed. Also, data on the Community and Supportive Services (CSS) component of the HOPE VI grants were not analyzed. Thus, there remain many important unanswered questions about the ultimate scope of mixed-income housing production through HOPE VI. This report provides a unique analysis of all 260 HOPE VI revitalization grants.
National Initiative on Mixed-Income Communities Formally launched at the Jack, Joseph and Morton Mandel School of Applied Social Sciences at Case Western Reserve University in 2013, the National Initiative on Mixed-Income Communities (NIMC) is a resource for research and information about mixed-income
1 Brophy and Smith, 1997; Cisneros and Engdahl, 2009; Joseph, 2006, 2013; Joseph, Chaskin, and Webber, 2007; Khadduri, 2001; Kleit, 2005; Popkin et al., 2004. 2 For background on the HOPE VI program see, for example, Cisneros and Engdahl, 2009; GAO, 1997; Fosburg et. al, 1996; Holin et al., 2003; Popkin et al., 2004; Popkin, 2010; Popkin et al. 2010; Turbov and Piper, 2005
National Initiative on Mixed-Income Communities 7 | P a g e Case Western Reserve University communities. Our mission is to help reduce urban poverty and promote successful mixed- income communities by facilitating high-quality research and making information and evidence easily available to policymakers and practitioners. To advance this mission, NIMC conducts research and evaluation, provides technical assistance and strategic consultation, compiles and disseminates data and literature on mixed-income developments and maintains networks among policymakers, practitioners and researchers.
NIMC researchers and collaborators have conducted a wide range of research and evaluation projects in the field of mixed-income development. This includes involvement in the national Choice Neighborhoods Initiative evaluation, the national Jobs Plus Pilot evaluation, a seven-year study of the Chicago Plan for Transformation, support for the HOPE SF evaluation in San Francisco, and evaluation of The Community Builders’ Cascade Village development in Akron, Ohio. The Chicago research has generated over fifteen academic articles and ten research briefs and a recently published book co-authored with Robert Chaskin of the University of Chicago Integrating the Inner City: The Promise and Perils of Mixed-Income Public Housing Transformation. In 2013, NIMC completed its first Scan of the Field on the topic of Social Dynamics in Mixed-Income Developments with information on 31 developments in the U.S. and Canada. In 2014, NIMC completed its second Scan of the Field on the topic of Resident Services in Mixed-Income Developments with information on 60 developments in the U.S. and Canada. Research scans, briefs and articles are available at nimc.case.edu.
Project Description This report provides a descriptive analysis of the quarterly report data from HOPE VI revitalization grants, 1993 through 2014. Main Street Grants, which were also awarded through the HOPE VI program, are excluded from this analysis as they focus on rejuvenating downtown business districts rather than on residential housing. These quarterly reports were retrieved from HUD in 2015 and provide the most comprehensive documentation available of units produced through the HOPE VI program.
The overarching research question that motivates this study is: What is the income and tenure mix of housing units that have been produced through the HOPE VI program? Income and tenure mix data provide insight into the nature of the mixed-income developments produced with HOPE VI funding and therefore the potential impact these developments had on residents and communities through the provision of a mix of subsidized and market-rate housing.
Other key research questions include: ● How does the production compare with what was agreed to by the housing authority and developer? ● How does unit production vary by factors such as region, size of grant and nature of the proposed income and tenure mix?
National Initiative on Mixed-Income Communities 8 | P a g e Case Western Reserve University ● What have been the timeframes of the various stages of production: relocation, demolition, construction, occupancy? ● What proportion of occupancy is re-occupancy by former residents? What is the return rate of original residents to the redeveloped properties? ● What were the sources of the redevelopment budgets and how much was leveraged from non-federal sources? ● What services were provided through the Community and Supportive Services (CSS) component of the HOPE VI program and how many residents were engaged? How many were placed and retained in jobs?
Although our analysis was primarily descriptive in nature, some of our key research questions have high policy relevance and are of interest to the field more broadly. For example, the question of whether HOPE VI grantees successfully produced the units they agreed to, especially public housing and affordable units, is of great concern for the preservation of affordable housing. The levels of income and tenure mix provide evidence about the extent of the deconcentration of poverty through the incorporation of market-rate rental and homeownership. A very important issue is the relocation of the original public housing residents and the extent of successful re-occupancy of former residents within new developments. This speaks to the question of forced relocation and who was able to benefit directly from living in the mixed- income redevelopment. Finally, understanding the timeframes of the stages of production, specifically the length of the often-extensive relocation and construction phases, help illuminate the practical challenges of the relocation and re-occupancy process.
Data from HUD program reports, which include projected and actual units produced, occupancy and re-occupancy, phases of construction, and financing were extracted and compiled for analysis. We acknowledge and thank Dr. Lawrence Vale and his team at MIT’s Resilient Cities and Housing Initiative for their collaboration in this endeavor. Separate quarterly and national reports on Community Supportive Services were extracted and compiled for analysis. Descriptive and comparative analyses of the 260 revitalization grantee reports and Community Supportive Services reports were completed using SPSS statistical software.3
Analyses in this report describe the data available in the HUD quarterly reports, which provide the most comprehensive look at the unit production and financial information of HOPE VI developments available to date. However, we have found inconsistencies when comparing these data to alternative data collected by NIMC through individual housing authorities and developers. Data collection efforts by the MIT team have yielded similar inconsistencies. The inconsistencies are greatest for projected unit data. The quarterly project reporting was used by HUD and by the grantees as an administrative tracking tool rather than a research tool and the projection numbers were updated over time, as new targets were agreed to between HUD and the
3 Significance testing was not conducted because we analyzed the total population of 260 revitalization grants.
National Initiative on Mixed-Income Communities 9 | P a g e Case Western Reserve University grantee. Thus, the projected numbers used in our analysis do not reflect the original number of units projected at the very beginning of each project, but instead represent the planned units eventually agreed to by HUD and the grantee. As a result, this analysis represents the difference between most recent agreed-upon projection numbers and actual production (as of the third quarter of 2014), not between original projection numbers and actual production.4
This report is structured as follows. First, we provide a summary of key findings. Next we describe our findings in greater detail describing the HOPE VI grant awards, unit production, income mix, tenure mix, projected and actual production, return rates, timeframes of production stages, re-occupancy, funding sources, and finally Community Supportive Services. We then discuss these findings and conclude by proposing implications for research and policy.
4 Efforts are currently underway, led by colleagues at MIT, to compile and analyze data on original unit production projections which will help fill this important knowledge gap.
National Initiative on Mixed-Income Communities 10 | P a g e Case Western Reserve University II. Summary of Key Findings
Below is a summary of key findings regarding redevelopment progress, financing and Community and Supportive Services (CSS).
Redevelopment Progress
Unit Production ● Overall, the HOPE VI redevelopment mainly produced replacement public housing units, with the additional inclusion primarily of “affordable” housing units (financed with low- income housing tax credits) and with limited inclusion of market-rate housing. Of the total 97,389 units produced thus far, 57% of units were for public housing, 30% were for affordable units and 13% were for market-rate units. ● Almost half of all sites (47%) have not built any market-rate units. ● Of the total units produced, 85% are rental units. About 10% of the public housing units were for homeownership. Over 40% of HOPE VI sites produced some of these public housing homeownership units. ● Older sites (12 years or longer) have a higher proportion of public housing units than newer sites. The newer sites averaged 54% public housing, 35% affordable units and 11% market-rate units, while the older sites averaged 68% public housing, 24% affordable units and 8% market-rate units. ● The greatest production of units occurred in the first ten years of the HOPE VI program. The greatest number of units was also demolished during this time. Starting in 2003, there was a leveling off of demolition, relocation and construction. ● Grantee award cohorts after 2003 did not produce nearly as much homeownership as those cohorts before 2003. ● The Southeast region had the largest number of constructed units. There were regional differences in the income mix. The New York-New Jersey region had the largest percentage of public housing units, the Northwest region had the largest percentage of affordable units and the Great Plains region had the largest percentage of market-rate units. ● We categorized different types of income mixes using Vale and Shamsuddin’s (2014) typology, which uses four categories of unit mix: Narrow Low-Income (public housing and affordable), Polarized Bi-Modal (public housing and market-rate), All But The Poorest (affordable and market-rate) and Broad Continuum (public housing, affordable and market-rate). Most redevelopments had a Narrow Low-Income mix (69.1%), followed by Broad Continuum (25.5%) and Polarized Bimodal (5%). Only one site had an All But The Poorest mix.
National Initiative on Mixed-Income Communities 11 | P a g e Case Western Reserve University ● Although the Narrow Low-Income sites produced more public housing replacement units, the Broad Continuum sites actually produced more total housing by adding more units of different income ranges. ● While HOPE VI production resulted in a decrease of public housing units, the redevelopments mainly produced subsidized housing: either public housing replacement units or “affordable” units financed with the low-income housing tax credit. The 98,592 demolished public housing units were replaced with 55,318 public housing units in addition to 28,979 affordable units, effectively replacing 85% of the original public housing units with units intended to be affordable to low-and moderate-income residents. ● Overall, 88.3% of the units that were projected (per the most recent agreements with HUD) have been actually produced. o A higher percentage of units that were projected have been produced for rental units (91.9%) than for homeownership units (72.7%). o A higher percentage of units that were projected have been produced for public housing units (94%) than affordable (87.1%) and market-rate units (71.8%). ● Overall, almost the same number of mixed-income units have been constructed to replace the 100% public housing units that have been demolished, 97,389 mixed-income units have been built to replace 98,592 public housing units. ● Through HOPE VI redevelopment, 43,274 units have been lost from the public housing stock.
Return and Relocation ● Of the 96,476 units that have been produced and occupied, only 19,993 units (20.7%) have been occupied by original tenants at each development. ● Of the 55,318 constructed units that are designated as replacement units for public housing residents, only 36.1% have been occupied by residents of the original development. These “re-occupancy” rates generally declined over time. ● Of the households originally relocated from the developments, only an average of 27.6% have returned to the new units (a median of 18.2%). ● Newer sites have thus far experienced lower return rates (21.1%) than older sites (33.8%), which may be mainly a function of time and a lag in construction and occupancy. But the return rate at the older sites likely indicates the upper end of the average that can ultimately be expected. ● Of the relocated residents, 5.5% were evicted and 11.9% were deceased or otherwise left the public housing development and did not qualify to return.
Timeframes of Redevelopment Progress ● As would be expected, given the variations in size of the sites, timeframes for relocation, demolition, construction and occupancy had large ranges, from weeks to years. ● The average duration of the relocation phase (694 days) was the longest followed by construction (667 days), demolition (516 days) and occupancy (260 days).
National Initiative on Mixed-Income Communities 12 | P a g e Case Western Reserve University ● Comparing sites of similar unit sizes, we found that there remained considerable variation in ranges. Among small and medium size sites, the longest phase was construction, but among large sites, the longest phase was relocation. Occupancy was the shortest phase regardless of site unit size.
Financing ● A total of $6B HOPE VI funds were expended. The average HOPE VI expended per site was $22.9M. ● The data document a decrease of HOPE VI funding over time with significant cuts in the program beginning in 2000. Newer sites received an average of $19.8M of HOPE VI funding and older sites received an average of $26.4M. ● The HOPE VI funds leveraged $11B additional public and private funds for the redevelopments for a total of $17B expended from all sources. The average total funding for each site was $65.3M. ● Sites with units targeting higher-income income residents consistently leveraged more funding. Narrow Low-Income sites (Mean total funding = $57M), which have only public housing and affordable units, generally had less funding and were not as able to leverage additional funds as much as projects that included market-rate or homeownership ((Broad Continuum [Mean total funding = $86.9M] and Polarized Bimodal [Mean total funding = $78.6M]).
Community and Supportive Services ● Services included employment support, job skills training programs, high school or equivalent education, English as a Second Language (ESL) courses, child care, transportation assistance, counseling programs, and substance abuse programs. Overall, most CSS programs exceeded most of their enrollment goals across services. ● The services with the highest number of enrollments were employment preparation/ placement/retention, transportation assistance, counseling programs, job skills training programs and childcare. ● There were differences in the number of services offered at sites by tenure mix, size and development age. o Larger sites had a higher average proportion (43.1%) of residents getting CSS support than smaller sites (28.7%). o Newer sites had a higher average proportion (44.3%) of residents getting CSS support than older sites (27.6%). ● Generally, while enrollment in the various programs and services was high, completion was much lower and often fell short of goals. o Employment: Placed more residents than the stated goals for total new job placements, however, the number of residents currently employed and the number employed for 6 months were far below their goals.
National Initiative on Mixed-Income Communities 13 | P a g e Case Western Reserve University o Economic Development: Enrollment goals for entrepreneurship training were exceeded, but the total number of residents who completed programs in this area was less than the goal. There were also fewer resident-owned businesses or residents employed by those businesses than hoped. o Homeownership: Residents successfully enrolled in homeownership counseling, but the number of residents who completed counseling and the number who purchased a home were less than expected. o Education: Generally, sites fell short of completion goals for high school or equivalent education services.
National Initiative on Mixed-Income Communities 14 | P a g e Case Western Reserve University III. Findings In this section we describe key findings from our analysis of HOPE VI revitalization grants awarded from 1993-2011 from information available as of the fourth quarter of 2014. The findings are organized as follows. First, we describe the grant awards given as part of the HOPE VI program at the national, regional and state levels. Second, we provide a descriptive and comparative analysis of unit production including income and tenure mix and redevelopment timeframes. Third, we provide a descriptive and comparative analysis of relocation and return. Last, we describe and compare enrollment and completion of activities in the Community and Supportive Services program.
Grant Awards
Nationwide Nationwide, a total of 260 HOPE VI revitalization implementation awards were granted to housing authorities. An additional 285 demolition-only grants were awarded but these are not included in this analysis. Of the 134 cities or counties that received revitalization implementation grants, there were 10 cities that received more than five grants: Chicago (nine grants), Atlanta, Baltimore, and Washington, DC (seven grants each), Milwaukee (six grants), Boston, Charlotte, Memphis, Philadelphia, and Seattle (five grants each). Figure 1 shows the number of grants by grantee award year. When we refer to award year data, this means the awards to a cohort of grantees in a particular year. An average of 14 awards were given in each of the grant years from 1993 to 2010 with the majority of grants given between 1995 and 2002. The highest number of grants in any single year was 28 in 1998. Funding for the HOPE VI program dramatically decreased after the 2003 award year due to the Bush administration’s efforts to eliminate the program. The funding by award year ranged from $197.1M to $1.7B with an average of $948.7M and a median of $1.1B.5
5 In the 2011 award year, only a single grant was awarded with $1.5M of total funding. This information was excluded from the award year analysis.
National Initiative on Mixed-Income Communities 15 | P a g e Case Western Reserve University Figure 1: Number of grants by award year
Financing Sources of financing for HOPE VI projects include HOPE VI grants, other public housing funding, other federal funding, and all other non-federal funding. A total of $6B in HOPE VI funds were expended. These HOPE VI funds leveraged $11B additional funds for the projects for a total of $17B expended from all sources to the 260 developments. For every one dollar of HOPE VI funds expended, about $1.8 dollars were leveraged for the projects. Specifically, $1.7B (10.08%) of other public housing funding, $1.5B (9.12%) of other federal funding, and $7.7B (45.62%) of all non-federal funding were expended.
In most cases projects have spent their grant agreement budgets with 85% of total budgeted funds spent. A total of 95.3% of the $6.3B HOPE VI funding budgeted has been actually spent.
Unit Production A total of 75,410 households have been relocated, 98,592 units have been demolished, and 97,389 units have been produced (this included 85,934 newly-constructed units and 11,455 units rehabbed units). Of the 96,476 units that have been produced and occupied, 19,993 units (20.7%) have been occupied by original tenants at each development. We’ll return later to the topic of the relocation and return of original tenants. Table 1 compares projected6 and actual construction.
6 Due to the administrative nature of the reports analyzed, projected unit data included in our analysis does not necessarily reflect the original projections included in grant proposals or original grant agreements. Instead, this can best be understood as the projected numbers in the most recent agreement between the grantee and HUD.
National Initiative on Mixed-Income Communities 16 | P a g e Case Western Reserve University Table 1: Comparison of projected and actual unit production
Projected Actual Public housing 53,226 49,949 93.8%
Affordable 26,674 23,899 89.6% Rental Market 9,786 8,530 87.2%
Total 89,686 82,378 91.9%
Public housing 5,594 5,369 96% Affordable 5,080 Home 6,607 76.9% Construction Owner Market 8,459 4,562 53.9%
Total 20,660 15,011 72.7%
Public housing 58,820 55,318 94% Affordable 28,979 Grand 33,281 87.1% Total Market 18,245 13,092 71.8%
Total 110,346 97,389 88.3%
As shown in Table 1, a total of 88.3% of the units projected in the grantee agreements were actually produced. Notably, 91.9% of the projected rental units were produced, while only 72.7% of the projected homeownership units were produced. Also, fewer actual market-rate units (71.8%) have been produced than public housing (94%) and affordable units (87.1%) in comparison to projected production. Barely over half (54%) of the project market-rate for-sale production has been completed, a clear sign of the impact of the 2008 housing market crash.
Table 2 presents information on relocation, occupancy and occupancy by original residents.
Table 2: Relocation, evictions, occupancy and re-occupancy
Households relocated 71,283 Evictions 4,127 Total Units Occupied 96,476 Units Occupied by 19,993 Returning Residents
National Initiative on Mixed-Income Communities 17 | P a g e Case Western Reserve University HOPE VI Grants by Region The map in Figure 2 shows the regions designated by the U.S. Department of Housing and Urban Development.
Figure 2: HUD’s Regions
Source: http://portal.hud.gov/hudportal/HUD%3Fsrc%3D/localoffices/regions
Tables 3 and 4 summarize financial information and unit construction information across HUD’s regions.7 Compared to other regions, the Southeast-Caribbean (Region 4) received the largest amounts of HOPE VI funding and produced the largest number of units. In comparison, the Rocky Mountain (Region 8) obtained the smallest amount of funding and constructed the smallest number of units. The HOPE VI funds allocated to the Southeast-Caribbean leveraged $2.7B additional funds for the projects for a total of $4.3B expended from all sources. In the Rocky Mountain region, for every one dollar of HOPE VI funds expended, about 3.3 dollars were leveraged, demonstrating the highest leverage ratio. On average, 26 grants were awarded per region.
7 Region 1 (3 out of 6 states): Connecticut, Vermont*, Massachusetts, Maine*, New Hampshire*, Rhode Island; Region 2 (2 of 2): New York, New Jersey; Region 3 (6 of 6): Pennsylvania, Virginia, West Virginia, Maryland, Delaware, District of Columbia; Region 4 (9 of 10): Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, Tennessee, Puerto Rico, U.S. Virgin Islands*; Region 5 (6 of 6): Illinois, Indiana, Michigan, Minnesota, Ohio, Wisconsin; Region 6 (3 of 5): Arkansas*, Louisiana, New Mexico*, Oklahoma, Texas; Region 7 (1 of 4): Kansas*, Iowa*, Missouri, Nebraska*; Region 8 (2 of 6): Colorado, Montana, North Dakota*, South Dakota*, Utah*, Wyoming*; Region 9 (2 of 4): California, Arizona, Hawaii*, Nevada*; Region 10 (2 of 4): Washington, Alaska*, Idaho*, Oregon * No implementation grant in these states National Initiative on Mixed-Income Communities 18 | P a g e Case Western Reserve University Table 3: Financial information by HUD regions
HOPE VI Total Leverage Name Number of Grants Expended Expended Ratio Southeast- Region 4 75 28.8% $1.6B $4.3B 1 : 1.7 Caribbean Region 3 Mid-Atlantic 43 16.5% $903.4M $2.5B 1 : 1.7
Region 5 Midwest 41 15.8% $1B $2.6B 1 : 1.6 New York-New Region 2 24 9.2% $586.7M $1.9B 1 : 2.2 Jersey Region 6 Southwest 19 7.3% $488.2M $1.2B 1 : 1.4
Region 9 Pacific 18 6.9% $391.5M $1.2B 1 : 2.1
Region 1 New England 15 5.8% $346.6M $1.2B 1 : 2.5
Region 10 Northwest 12 4.6% $323.9M $1.3B 1 : 2.9
Region 7 Great Plains 8 3.1% $179.1M $420.8M 1 : 1.3
Region 8 Rocky Mountain 5 1.9% $87.8M $380.4M 1 : 3.3
Total - 260 100% $6B $17B - Note: In order of number of grants
National Initiative on Mixed-Income Communities 19 | P a g e Case Western Reserve University Table 4: Unit construction information by HUD regions
Income Mix Tenure Mix Public Home Total Affordable Market Rental housing Owner Southeast- 15,997 4,607 11,088 (35%) 4,730 (15%) 27,208 (86%) 31,815 Caribbean (50%) (14%) 3,561 Mid-Atlantic 8,115 (64%) 3,600 (28%) 997 (8%) 9,151 (72%) 12,712 (28%) 1,932 Midwest 8,419 (59%) 3,378 (23%) 2,592 (18%) 12,457 (87%) 14,389 (13%) New York- 1,052 6,424 (69%) 2,705 (29%) 179 (2%) 8,256 (89%) 9,308 New Jersey (11%) 749 Southwest 5,162 (66%) 1,587 (20%) 1,018 (13%) 7,018 (90%) 7,767 (10%) 644 Pacific 3,340 (59%) 1,846 (33%) 479 (8%) 5,021 (89%) 5,665 (11%) 387 New England 2,701 (67%) 844 (21%) 515 (13%) 3,673 (90%) 4,060 (10%) 1,424 Northwest 2,429 (37%) 2,881 (43%) 1,344 (20%) 5,230 (79%) 6,654 (21%) 514 Great Plains 1,656 (51%) 690 (21%) 933 (28%) 2,765 (84%) 3,279 (16%) Rocky 1,075 (62%) 360 (21%) 305 (18%) 1,599 (92%) 141 (8%) 1,740 Mountain Note: In order of number of grants
Although the Southeast-Caribbean (Region 4) had the largest number of grants and constructed the largest number of units, the New York-New Jersey (Region 2) had the largest percentage of public housing, the Northwest (Region 10) had the largest percentage of affordable housing and the Great Plains (Region 7) had the largest percentage of market-rate housing.
HOPE VI Grants by State Of the 36 states that received HOPE VI implementation grants, there were 10 states that received more than 11 grants: Pennsylvania (18), New Jersey (17), Florida and North Carolina (14), Illinois and Texas (13), California, Ohio, and Tennessee (12 each), Georgia (11). Table 5 and 6 summarize the financial and unit production information by states.
National Initiative on Mixed-Income Communities 20 | P a g e Case Western Reserve University Table 5: Implementation grant information by states
Number of HOPE VI Total expended State Budgeted grants Expended (All sources) New Jersey 17 $435.3M $405M $1.4B Illinois 13 $367.6M $352.8M $1.3B
Pennsylvania 18 $367.7M $367.7M $1.2B Washington 9 $255.6M $254.4M $1B
Georgia 11 $279.9M $279.2M $994.9M California 12 $324.8M $318.7M $799M
North Carolina 14 $347M $324.4M $777.5M Massachusetts 9 $203.1M $195.3M $730.1M
Kentucky 8 $166.7M $143.2M $703.1M Texas 13 $373.5M $362.9M $662.8M
Ohio 12 $326.2M $305.7M $640.6M Florida 14 $275.6M $268.2M $606M
Tennessee 12 $301.7M $289.7M $539.6M District of 7 $181.2M $181.2M $486.5M Columbia New York 7 $197.1M $181.7M $473.9M Louisiana 5 $104.1M $96.6M $470.1M
Maryland 9 $209.7M $182.9M $469M Missouri 8 $180.2M $179.1M $420.8M
Arizona 6 $100.9M $72.8M $420.3M South Carolina 7 $142.6M $142.6M $394.3M
Connecticut 5 $131.4M $131.4M $393.2M Colorado 4 $94.2M $86.8M $379.5M
Virginia 7 $150.2M $137.7M $296.8M Alabama 6 $119.7M $119.7M $248.4M
Michigan 4 $128.7M $116.1M $247.8M Oregon 3 $70.5M $69.5M $240.4M
Wisconsin 6 $136.5M $136.5M $217.4M
National Initiative on Mixed-Income Communities 21 | P a g e Case Western Reserve University Number of HOPE VI Total expended State Budgeted grants Expended (All sources) Minnesota 2 $34.2M $32M $105.9M Indiana 4 $78.8M $77.6M $100.5M
Rhode Island 1 $20M $20M $74M Puerto Rico 1 $50M $39.7M $62.8M
Oklahoma 1 $28.6M $28.6M $54.6M West Virginia 1 $17.1M $17.1M $31.1M
Mississippi 2 $52.3M $29.6M $30.2M Delaware 1 $16.8M $16.8M $28.8M
Montana 1 $0.9M $0.9M $0.9M
It is notable that New Jersey, Pennsylvania, Illinois and Washington expended the most funds.
Table 6: Unit production information by states
Public housing Affordable Market State Total units (%) (%) (%) North Carolina 2,752 (38.8%) 3,362 (47.4%) 984 (13.9%) 7,098 Georgia 2,326 (35.1%) 2,456 (37.1%) 1,843 (27.8%) 6,625
Illinois 2,672 (43.3%) 1,734 (28.1%) 1,762 (28.6%) 6,168 New Jersey 3,705 (61.2%) 2,245 (37.1%) 106 (1.8%) 6,056
Texas 4,000 (72.1%) 1,121 (20.2%) 424 (7.6%) 5,545 Washington 1,923 (34.7%) 2,499 (45.1%) 1,114 (20.1%) 5,536
Pennsylvania 3,916 (77.1%) 756 (14.9%) 408 (8%) 5,080 Florida 2,878 (58.3%) 1,665 (33.7%) 396 (8%) 4,939
Kentucky 2,565 (60.5%) 1,079 (25.4%) 599 (14.1%) 4,243 Ohio 2,849 (72.3%) 669 (17%) 423 (10.7%) 3,941
California 2,442 (63.1%) 1,309 (33.9%) 116 (3%) 3,867 Tennessee 2,351 (62.5%) 885 (23.5%) 527 (14%) 3,763
Missouri 1,656 (50.5%) 690 (21%) 933 (28.5%) 3,279
National Initiative on Mixed-Income Communities 22 | P a g e Case Western Reserve University Public housing Affordable Market State Total units (%) (%) (%) New York 2,719 (83.6%) 460 (14.1%) 73 (2.2%) 3,252 South Carolina 1,261 (46.5%) 1,220 (45%) 229 (8.5%) 2,710
District of Columbia 1,094 (43.6%) 1,099 (43.8%) 316 (12.6%) 2,509 Virginia 1,568 (62.6%) 752 (30%) 186 (7.4%) 2,506
Massachusetts 1,776 (72.4%) 362 (14.8%) 315 (12.8%) 2,453 Maryland 1,305 (55.9%) 946 (40.5%) 84 (3.6%) 2,335
Louisiana 875 (48.2%) 356 (19.6%) 586 (32.3%) 1,817 Arizona 898 (49.9%) 537 (29.9%) 363 (20.2%) 1,798
Colorado 1,061 (61.5%) 360 (20.9%) 305 (17.7%) 1,726 Michigan 1,042 (61.9%) 477 (28.3%) 165 (9.8%) 1,684
Alabama 1,219 (72.9%) 302 (18.1%) 152 (9.1%) 1,673 Connecticut 817 (60.7%) 369 (27.4%) 159 (11.8%) 1,345
Wisconsin 1,173 (88.2%) 131 (9.8%) 26 (2%) 1,330 Oregon 506 (45.3%) 382 (34.2%) 230 (20.6%) 1,118
Indiana 471 (58.9%) 240 (30%) 88 (11%) 799 Mississippi 425 (78.1%) 119 (21.9%) 0 (0%) 544
Minnesota 212 (45.4%) 127 (27.2%) 128 (27.4%) 467 Oklahoma 287 (70.9%) 110 (27.2%) 8 (2%) 405
Rhode Island 108 (41.2%) 113 (43.1%) 41 (15.6%) 262 Puerto Rico 220 (100%) 0 (0%) 0 (0%) 220
Delaware 160 (100%) 0 (0%) 0 (0%) 160 West Virginia 72 (59%) 47 (38.5%) 3 (2.5%) 122
Montana 14 (100%) 0 (0%) 0 (0%) 14
In all states the greatest number of units have been produced for replacement public housing, followed by affordable and market-rate housing. Delaware, Puerto Rico and Montana only produced public housing. New York, Pennsylvania, Mississippi and Wisconsin produced over
National Initiative on Mixed-Income Communities 23 | P a g e Case Western Reserve University 75% public housing. Meanwhile, New York, Maryland, West Virginia, Wisconsin, Oklahoma and California all produced less than four percent market-rate housing.
Redevelopment Progress
Redevelopment Progress by Award Year In this section we describe redevelopment progress in terms of demolition, relocation and construction. A total of 75,410 households have been relocated, 98,592 units have been demolished and 97,389 units have been constructed as part of the HOPE VI program. Figure 3 below shows redevelopment progress by award year for relocated households and demolished and constructed units from 1993-2010.8 Figure 4 shows cumulative redevelopment progress by award year. The greatest increase in production occurred between the 1993 and 1996 award years. Cohorts 1996 and 1999 demolished the most units. Cohorts 1998, 2000 and 2002 have produced the most units with a significant drop in the pace of production following the 2003 award year, followed by another peak in 2005.
Figure 3: Redevelopment progress by award year
8 In the award year information of 2011, only 12 units were constructed. This information was excluded.
National Initiative on Mixed-Income Communities 24 | P a g e Case Western Reserve University Figure 4: Cumulative redevelopment progress by award year
100000
90000
80000
70000
60000
50000
40000
30000
20000
10000
0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Cumulative Construction Cumulative Relocation Cumulative Demolition
Overall, Figure 4 indicates that there were basically two different redevelopment phases of the HOPE VI program. There was the phase between 1993 and 2003 where production increased steadily. Then in the second phase between 2004 and 2010, there was a relative leveling off of production. Prior to 2003, the average number of relocated households per year was about 6,192 units. From 2004 to 2010, the average number of households relocated per year declined to about 1,043 units. Similarly, the total number of demolished units dropped between 2003 and 2004. On average, about 8,033 units were demolished annually between 1993 and 2003, while about 1,467 units were demolished annually between 2004 and 2010.
Between 1993 and 2010 the total number of constructed units was 97,389 units. Like the numbers of relocated households and demolished units, the total pace of unit construction dramatically decreased between 2003 and 2004. During the period from 1993 to 2003, the average number of units produced each year was about 7,788 units. During the period from 2004 to 2010, however, the average number of units produced annually decreased to about 1,733 units. The highest unit production was 8,953 units in 1998, while the lowest unit production was 777 units in 2006.
National Initiative on Mixed-Income Communities 25 | P a g e Case Western Reserve University Figure 5 below shows cumulative demolition of public housing units and the relative construction of public housing, affordable and market-rate units by grantee award year.
Figure 5: Cumulative demolition and three types of construction by award year
100000
90000
80000
70000
60000
50000
40000
30000
20000
10000
0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
PH Construction AFF Construction MKT Construction Cumulative Demolition
Income/Tenure Mix The mix of housing types within each development is displayed in Figures 6 and 7. Figure 6 shows the income mix across the 2599 developments in three categories: public housing, “affordable” and market-rate, without distinguishing between rental and for-sale units. As described earlier, “affordable” units are an additional tier of subsidized units, primarily funded by the low-income housing tax credit program which usually house tenants who are low-income but not at the poverty levels of public housing residents. Each bar displayed represents the mix present at an individual housing development. Figure 7 highlights tenure mix across the 259 developments with just two categories: rental and homeownership.
9 Among 260 sites there is one site that did not produce any units. This site had planned to produce 100 units, but not produced any as of the end of the reporting period. Thus, for all analysis using unit production information, a total of 259 sites were analyzed excluding the site.
National Initiative on Mixed-Income Communities 26 | P a g e Case Western Reserve University
Figure 6: Income mix across 259 HOPE VI developments (Each bar represents a single actual development)
National Initiative on Mixed-Income Communities 27 | P a g e Case Western Reserve University Figure 7: Tenure mix across 259 HOPE VI developments (Each bar represents a single actual development)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Rental For Sale
National Initiative on Mixed-Income Communities 28 | P a g e Case Western Reserve University
Unit Production by Subsidy Type As illustrated in Table 1, a total of 97,389 units have been produced at HOPE VI sites.
Public Housing Unit Production Among the total production of 97,389 units, 55,318 units (56.8%) were for public housing. Among those public housing units, 49,949 units (90.3%) were public housing rental units and 5,369 units (9.7%) were public housing homeownership. It is notable that about 10% of public housing units across sites were for homeownership. Section 32 of the U.S. Housing Act of 1937 outlines three ways public housing authorities can help public housing residents purchase homes: (1) set aside public housing units for purchase by low-income residents, (2) provide funds to public housing residents to help them purchase homes, or (3) housing authorities can buy homes for the purpose of selling them to low-income families.10 Among the 259 sites, a relatively high number, 105 sites (40.5%), produced public housing for-sale units. There were 34 sites which produced 50 or more public housing homeownership units.
Affordable Housing Unit Production Approximately 30% of units produced (28,979 units) were for “affordable housing.” Of these affordable units, 23,899 (82.5%) were rental units and 5,080 (17.5%) were for-sale units. There were 40 sites (15.4%) which did not produce any of these middle tier affordable units to supplement the replacement public housing that they produced.
Market-Rate Housing Unit Production Approximately 13% of units produced (13,092 units) were market-rate units, with 8,530 (65.2%) rental units and 4,562 (34.8%) were for-sale units. There were 121 sites (46.7%) which did not produce any market-rate units.
Tenure Mix Figure 8 below shows unit production by award year for rental and homeownership from 1993- 2010. Both rental and homeownership unit production declined between 2003 and 2004. During the period from 1993 to 2003, the average number of rental units produced annually was 6,509 units and the number of homeownership units produced was 1,280. During the period from 2004 to 2010, the average number of rental units produced annually was about 1,600 units and the number of homeownership units produced was 133.
10 For more information about public housing homeownership reference “Guidance for PHAs Developing a Section 32 Homeownership Plan”, U.S. Department of Housing and Urban Development, 2003 http://portal.hud.gov/hudportal/HUD?src=/program_offices/public_indian_housing/centers/sac/homeownership National Initiative on Mixed-Income Communities 29 | P a g e Case Western Reserve University
Figure 8: Rental and homeownership unit production by award year
In sum, among the total 97,389 units produced, 82,378 units (84.6%) were for rental housing and 15,011 units (15.4%) were for homeownership.
Projected versus Actual Production
Difference between Total Units Overall, a total of 12,857 of the units that were projected in the most recent grant agreements have not yet been produced. Among the 259 sites, 66 sites (25.5%) produced less than their most recent projections while 193 sites (74.5%) produced units as planned. No site produced more units than projected. On average, the HOPE VI sites produced 7.1% fewer units than their most recent projections. Thirty-two sites produced over 20% fewer units than planned, with the highest shortfall being 88.1%.
Difference between Public Housing Units A total of 3,402 units of public housing were not produced as planned. On average, the HOPE VI sites actually constructed just 13 (5.3%) fewer public housing units than projected. In all, the majority of sites (n = 218, 84.2%) have produced all the public housing units in their most recent agreements. Only 40 sites (15.8%) have produced fewer than projected. Twenty-eight sites produced over 20% fewer public housing units than planned. One site did not produce any of the projected public housing units. Of the 106 sites which produced public housing for-sale units, 87.7% (93 sites) produced units as planned and 12.3% (13 sites) produced less than projected.
National Initiative on Mixed-Income Communities 30 | P a g e Case Western Reserve University
Difference between Affordable Units A total of 4,302 units of affordable housing were not produced as planned. On average, these sites actually constructed 13 (7.3%) fewer affordable units than their projected units (n = 219). In all, the majority of sites (166, 75.8%) produced units as planned. Only 53 sites (24.2%) produced fewer than projected. Thirty sites produced over 20% fewer affordable units than their plans with the highest shortfall being 88.4%.
Difference between Market-Rate Units Market-rate production fell short of the planned production by 5,153 units. On average, these sites actually constructed 36 (14.7%) fewer market-rate units than projected (n = 143). In all, the majority of sites (n = 109, 76.2%) produced units as planned. Only 34 sites (23.8%) produced fewer than projected. Thirty-one sites produced over 20% fewer market-rate units than their plans. Five sites did not produce any of their projected market-rate units at all.
Difference between Rental and Homeownership Units For rental units, a total of 7,208 units were not produced as planned. On average, these HOPE VI sites produced 5.6% fewer rental units than their projected units. However, the majority of sites (n = 221, 85.7%) produced units as planned. Only 37 sites (14.3%) produced fewer than projected. Twenty-nine sites produced over 20% fewer rental units than their plans with the highest shortfall being 90.7%.
For homeowner units, a total of 5,649 units were not produced as planned. On average, these HOPE VI sites produced 14.6% fewer homeownership units than their projected units. A majority of sites produced units as planned (n = 143, 73%), while only 53 sites (27%) produced fewer than projected. Forty-four sites produced over 20% fewer homeownership units than their plans. Twelve sites did not produce any of their planned homeownership units at all.
National Initiative on Mixed-Income Communities 31 | P a g e Case Western Reserve University
Timeframes of Redevelopment Progress The HOPE VI project reports provided redevelopment progress timeframe information in addition to unit production information. We identified and analyzed the timeframes of the various stages of HOPE VI project production. The four stages are relocation, demolition, construction, and occupancy. Durations of the timeframes were calculated by identifying the earliest and latest start and finish dates for each stage of production.11
Each site had multiple phases for relocation, demolition, construction and occupancy. We calculated the number of days for the shortest and the longest phases within each site. We also calculated the average duration of each stage regardless of the separate phases within each site.
Relocation is the process of moving residents off-site prior to the demolition of the public housing units. Residents who are relocated may be relocated permanently or on a temporary basis until the new development is complete and ready for occupancy. Occupancy is the process of moving residents back into the newly constructed development.
To illustrate the different phases of development we outlined the phases for the Maverick Landing development in Boston, MA below. There were three redevelopment phases for this development:
In the first redevelopment phase, relocation (the process of moving households off-site) started in December 2002 and ended in May 2003, demolition started in June 2003 and ended in September 2003, construction started in November 2003 and ended in December 2004, and the first units were available for occupancy in the middle of December 2004 and all units were available by the end of that month.
In the second redevelopment phase, relocation started in June 2003 and ended in March 2005, demolition started in February 2005 and ended in August 2005, construction started in May 2005 and ended in October 2006, and in October 2006 these units were fully available.
In the last redevelopment phase, construction started in November 2003 and ended in March 2005, and in March 2005 these units were fully available.
11 Among the 260 sites, 234 sites provided relocation timeframes, 218 sites provided demolition timeframes, 215 sites provided construction timeframes and 206 sites provided occupancy timeframes National Initiative on Mixed-Income Communities 32 | P a g e Case Western Reserve University
Tables 7 to 10 show the shortest and longest phases and average durations of relocation, demolition, construction and occupancy.
Table 7: Relocation timeframes (days; N = 234)
M Median SD Min if > 0 Max
Shortest Phase 580 343 601 16 3,101
Longest Phase 813 571 710 33 5,079
Average Duration 694 466 623 33 3,643
Table 8: Demolition timeframes (N = 218)
M Median SD Min if > 0 Max
Shortest Phase 406 179 621 1 4,318
Longest Phase 627 322 812 23 5,204
Average Duration 516 268 677 23 4,318
Table 9: Construction timeframes (N = 216)
M Median SD Min if > 0 Max
Shortest Phase 425 365 266 7 1,995
Longest Phase 1,031 838 650 263 4,912
Average Duration 667 580 336 208 2,949
Table 10: Occupancy timeframes (N = 206)
M Median SD Min > 0 Max
Shortest Phase 143 76 233 1 1,438
Longest Phase 394 269 438 13 3,939
Average Duration 260 172 367 13 3,939
National Initiative on Mixed-Income Communities 33 | P a g e Case Western Reserve University
Comparative Analysis of Redevelopment Progress Thus far we have provided a basic descriptive analysis of the HOPE VI unit production, we now turn to a comparative analysis of these data. We compare unit production by region, income and tenure mix, size, and age of HOPE VI sites.
Region We begin our comparative analysis by analyzing production by region to understand how units have been produced in different areas of the country. As shown in Figure 2, HUD has designated ten regions across the United States.
Unit Production by Region On average, a total of 376 units were produced by sites in each region, and 56.9% of them were public housing units (Mean = 214 units). There were notable differences in the average site production by region. The Southeast-Caribbean (Mean = 424) produced more units per site than New England (Mean = 271) and the Mid-Atlantic (Mean = 303).
Table 11 shows the percentage of the three types of housing by region. Proportions of average public housing units were between 48.2% and 70.6%, the average affordable units were between 15.7% and 39.8%, and the average market-rate units were between 3.8% and 23.3%.
Table 11: Percentage of unit production by region (N = 259)
Public housing Market (%) Affordable (%) (%) M (SD) M (SD) M (SD) Rocky Mountain 70.6 (27.1) 15.7 (14.7) 13.7 (12.9)
Southwest 68.7 (28.7) 19.4 (22.3) 11.9 (21.8)
Mid-Atlantic 68.1 (26.1) 25.6 (25.1) 6.3 (10.1)
New England 64.2 (17.5) 25.3 (18.3) 10.5 (8.7)
Midwest 62.7 (23.9) 25.4 (17.4) 11.9 (14.6)
Pacific 58.6 (22.3) 32.5 (22.2) 8.9 (15)
New York-New Jersey 58 (27) 38.3 (24.9) 3.8 (10.5)
Southeast-Caribbean 56.3 (21.7) 32.6 (19.1) 11.1 (13.1)
Great Plains 50.1 (15.3) 26.6 (12.8) 23.3 (16.6)
Northwest 48.2 (20.5) 39.8 (16) 12 (15.4)
Total 60.6 (24) 29.3 (21.1) 10.1 (13.9)
Note: In order of public housing unit proportions National Initiative on Mixed-Income Communities 34 | P a g e Case Western Reserve University
Income/Tenure Mix
Income Mix In order to compare income and tenure mix of HOPE VI projects we’ve adopted the four- category typology of income mixing developed by Larry Vale and Shomon Shamsuddin (2014). This typology for describing the income mix was operationalized as a variable that could be used in the bivariate analyses. This variable has the following categories: (a) Narrow Low-Income, (b) Polarized Bimodal, (c) All But The Poorest, and (d) Broad Continuum. In our adaptation of this typology, the Narrow Low-Income type included developments with mostly public housing and affordable units. The Polarized Bimodal type included developments where there were primarily public housing and market-rate units and very few affordable units. The All But The Poorest type included developments that had no public housing units. The Broad Continuum type included developments that included a mix of public housing, affordable housing, and market-rate housing.
After reviewing the distribution of income mix among HOPE VI developments, we assigned each development to a category based on the following criteria. Narrow Low-Income developments have no more than 15% market-rate units. Polarized Bimodal developments have no more than 15% affordable units. All But The Poorest developments have no public housing units. Broad Continuum developments have at least 15% of all three unit types: public housing, affordable, and market-rate.
Figure 9: Count and proportion of mix type across developments (N=259)
66. 25.5%
Narrow Low-Income Polarized Bimodal 1. 0.4% All But The Poorest 13. 5.0% Broad Continuum
179. 69.1%
National Initiative on Mixed-Income Communities 35 | P a g e Case Western Reserve University
One hundred and seventy nine developments, about 70%, were in the Narrow Low-Income category, followed by Broad Continuum (25.5%) and Polarized Bimodal (5%). There was only one development in the All But The Poorest category.
Unit Production by Income Mix Table 12 shows the differences in average unit production by income mix. More units were produced on average in Broad Continuum sites than in Narrow Low-Income sites. More affordable units were produced on average in Broad Continuum sites than in Narrow Low- Income and Polarized Bimodal sites. More market-rate units were produced on average in Polarized Bimodal sites than in Broad Continuum or Narrow Low-Income sites.
Table 12: Comparison of average unit production by income mix (N = 258)
Public Market housing Affordable Total
n M (SD) M (SD) M (SD) M (SD) Narrow Low-Income 179 217 (184) 101 (110) 8 (20) 326 (212)
Broad Continuum 66 195 (103) 158 (122) 140 (112) 492 (285)
Polarized Bimodal 13 283 (173) 37 (54) 179 (223) 500 (391)
Total 258 214 (167) 112 (115) 51 (99) 377 (255)
Table 13 shows the difference in the percentage of unit production by income mix. The Broad Continuum developments had approximately 42% of public housing, 32% of affordable, and 26% of market-rate units on average, while the Narrow Low-Income developments had less than three percent of market-rate units. The Polarized Bimodal developments averaged 61% public housing and 34% market-rate housing.
Table 13: Comparison of average percentage of actual unit production by income mix (N = 258)
Public housing Affordable (%) Market (%) (%) M (SD) M (SD) M (SD) Narrow Low-Income 67.8 (23.8) 30 (23.4) 2.2 (4.1)
Broad Continuum 42.1 (11.6) 31.6 (11.5) 26.3 (10.3)
Polarized Bimodal 60.9 (16.6) 5.4 (5.7) 33.7 (15)
Total 60.9 (23.7) 29.2 (21.1) 10 (13.7)
National Initiative on Mixed-Income Communities 36 | P a g e Case Western Reserve University
Tenure Mix To compare the tenure mix of HOPE VI projects, the tenure mix was operationalized as a binary variable for use in bivariate analyses. This variable had the following categories: (a) Sites without any homeownership units (n = 75; 29%) and (b) Mixed-tenure sites having both rental and homeownership units (n = 184; 71%).
Unit Production by Tenure Mix Notable differences were identified for total unit production, affordable units, and market-rate units by tenure mix (Table 14). Mixed-tenure sites produced more units than sites without homeownership units.
Table 14: Comparison of average unit production by tenure mix (N = 259)
Public Market housing Affordable Total
n M (SD) M (SD) M (SD) M (SD) Rental-Only 75 197 (137) 87 (104) 28 (56) 312 (196) Developments Mixed-Tenure 184 220 (178) 122 (118) 60 (111) 402 (272) Developments
Table 15 shows the difference in the percentage of unit production by tenure mix. The sites without any homeownership units had approximately 66% public housing, 27% affordable, and 7% market-rate units on average, and the mixed-tenure sites had approximately 58% public housing, 30% affordable, and 11% market-rate units on average.
Table 15: Comparison of average percentage of actual unit production by tenure mix (N = 259)
Public housing Affordable (%) Market (%) (%) M (SD) M (SD) M (SD) Rental-Only 66 (25.1) 27.1 (22.3) 7 (10.4) Developments Mixed Tenure 58.5 (23.2) 30.2 (20.6) 11.4 (14.9) Developments
Comparison of Projected and Actual Unit Production by Tenure Mix Table 16 shows the average percent deficit of actual and projected unit production by tenure mix. The greatest deficit was in the production of market-rate units, followed by affordable and public housing units in both developments with and without homeownership.
National Initiative on Mixed-Income Communities 37 | P a g e Case Western Reserve University
Table 16: Average percentage deficit of production by tenure mix (N = 259)
Public Housing Affordable Market M (SD) M (SD) M (SD) Rental-Only -8.9 (20.7) -13.7 (22.9) -30.6 (38.6) Developments Mixed Tenure -3.8 (13) -5.3 (13.8) -9.9 (25.4) Developments
Size In addition to comparing sites by income and tenure mix, we also compare sites by unit size. The size variable was operationalized using the median value (319 units) of actual total unit production and dividing the 259 developments into either larger (having equal or more than 319 units) or smaller (having less than 319 units).
On average, the larger sites (n = 130) planned to construct 607 units and the smaller sites (n = 129) planned to construct 243 units. Table 17 illustrates the differences in unit production by size across housing types.
Table 17: Comparison of average unit production by size (N = 259)
Public Affordable Market Total housing n M (SD) M (SD) M (SD) M (SD) Larger (≥ 319 units) 130 302 (189) 165 (134) 87 (128) 554 (246)
Smaller (< 319 units) 129 124 (63) 58 (52) 14 (25) 197 (74)
Table 18 shows the differences in the percentage of unit production by size. On average, the larger sites had 56.9% of public housing and 12.8% of market-rate units, while the smaller sites had 64.4% of public housing and 7.4% of market-rate units. Larger sites tended to have a lower proportion of public housing and higher proportion of market-rate housing than smaller sites.
Table 18: Comparison of average percentage of actual unit production by size (N = 259)
Public housing Market (%) Affordable (%) (%) M (SD) M (SD) M (SD) Larger (≥ 319 units) 56.9 (24.6) 30.3 (21.6) 12.8 (14.9)
Smaller (< 319 units) 64.4 (22.8) 28.3 (20.6) 7.4 (12.3)
National Initiative on Mixed-Income Communities 38 | P a g e Case Western Reserve University
In addition, we analyzed the percentages of income mix by size. Among larger sites, 60% were Narrow Low-Income sites, while among smaller sites 78.3% were Narrow Low-Income. Among larger sites, 33.8% were Broad Continuum while among smaller sites 17.1% were Broad Continuum.
Age In addition to income and tenure mix and size, we also compared developments by age. A continuous age variable was calculated using the period between the dates of first unit availability for occupancy of each HOPE VI development and June 1st of 2015, when analysis of the data began.
For bivariate analysis, a categorical age variable was operationalized using the median value (11 years) of the age variable: older sites (aged equal or more than 12 years) and newer sites (aged equal or less than 11 years).
Table 19 illustrates the differences in unit production by age. There were notable differences in public housing, affordable and total unit production by age. Older sites constructed more public housing and total units, and less affordable units than newer sites.
Table 19: Comparison of average unit production by age (N = 259)
Public Affordable Market Total housing N M (SD) M (SD) M (SD) M (SD) Older sites (≥ 12 years) 127 267 (205) 96 (109) 53 (112) 417 (286) Newer sites (≤ 11 131 162 (94) 128 (119) 48 (86) 338 (216) years)
Age of the HOPE VI developments was associated with the proportion of public housing and affordable units. Newer sites tended to have a higher proportion of affordable and lower proportion of public housing units than older sites. Table 20 shows the differences in the percentage of actual unit production by age. On average, the older sites had 67.5% public housing and 23.5% affordable units, while the newer sites had 53.9% public housing and 34.9% affordable units.
National Initiative on Mixed-Income Communities 39 | P a g e Case Western Reserve University
Table 20: Comparison of average percentage of unit production by age (N = 259)
Public housing Affordable (%) Market (%) (%) M (SD) M (SD) M (SD) Older sites (≥ 12 67.5 (23.7) 23.5 (20) 9 (13.1) years) Newer sites (≤ 11 53.9 (22.5) 34.9 (20.7) 11.2 (14.7) years)
In addition, we analyzed the percentages of income mix by age. Among newer sites, 64.9% were Narrow Low-Income and among older sites 74% were Narrow Low-Income. Among newer sites, 30.5% were Broad Continuum while among old sites 19.7% were Broad Continuum.
Return Rates In this section we seek to answer two research questions: What percentage of relocated residents returned to “re-occupy” the redeveloped units? How does that vary by characteristics of the sites such as region, income and tenure mix, size and age? (Of the 259 developments, 23 developments did not provide data on re-occupancy.)
Return rates were calculated by dividing the number of units re-occupied by original residents by the number of households that were relocated. The return rates ranged from 0% to 100% with an average of 27.6% and a median of 18.2%. The relatively substantial difference between the average and the median is because the distribution is skewed with the bulk of sites in the low range and then a long tail of sites with higher return rates. However, when the 32 developments that were 100% public housing replacement are removed from the analysis, the average and median return rates for the developments that are mixed-income do not change very much at all. This is an indication that even in the sites that were 100% public housing replacement, the return rates of original residents were relatively low (33%).
Return Rates by Award Year Figure 10 below shows return rates by grantee cohort award year. The average return rates each year ranged from 6.3% to 44.7%. Return rates mostly declined between 1993 and 2000 cohorts, with a couple peaks in the 2001 and 2004 years followed by a sharp drop off in 2005 and another relatively small peak in 2008. These peaks and valleys follow a similar pattern to those of relocation, demolition, construction and occupancy displayed in Figure 1. Of course, a lower return rate over time is to be expected, particularly closest to the present time, since there has been less time for units to be completed and occupied. But even grants awarded ten years ago are showing extremely low return rates.
National Initiative on Mixed-Income Communities 40 | P a g e Case Western Reserve University
Figure 10: Return rates by award year
0.5
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
National Initiative on Mixed-Income Communities 41 | P a g e Case Western Reserve University
Comparative Analysis of Return Rates In this section we compare the return rates by region, income and tenure mix, size and age.
Region Table 21 compares return rates by region.
Table 21: Comparison of average rates by region (N = 236)
Return Rate (%)
n M (SD) New England 15 54 (31.1)
Great Plains 6 47.3 (35.1)
Pacific 16 33.9 (29.9)
New York-New Jersey 21 33.1 (25.4)
Rocky Mountain 4 31.8 (46)
Northwest 10 31.6 (16.6)
Mid-Atlantic 38 31.1 (26.1)
Southwest 15 30.2 (24.8)
Midwest 38 27.6 (21.9)
Southeast-Caribbean 73 14.4 (15.4)
Total 236 27.6 (25.4) Note: In order by return rate
Income/Tenure Mix
Return Rates by Income Mix We also compared return rates by income and tenure mix. To compare return rates by income mix we again used the Vale and Shamsuddin (2014) typology. Return rates were highest in the Polarized-Bimodal sites followed by Narrow Low-Income and Broad Continuum sites.
National Initiative on Mixed-Income Communities 42 | P a g e Case Western Reserve University
Table 22: Comparison of return rates by income mix (N = 235)
Return Rate (%)12 n M (SD) Narrow Low-Income 160 28.2 (27.1)
Broad Continuum 62 25.4 (19.9)
Polarized Bimodal 13 32.2 (27.4)
Total 235 27.7 (25.4)
Return Rates by Tenure Mix We also compared return rates by tenure mix. Sites without homeownership units had higher return rates compared to those with homeownership units.
Table 23: Comparison of return rates by tenure mix (N = 236)
Return Rate (%) n M (SD) Rental-Only 64 37.4 (29.3) Developments
Mixed Tenure 172 23.9 (22.8) Developments
Size In addition to income and tenure mix we also compared return rates by size. Return rates were extremely similar for both larger and small sites as shown in Table 24.
Table 24: Comparison of return rates by size (N = 236)
Return Rate (%) n M (SD) Larger (≥ 319 units) 125 27.6 (24.5)
Smaller (< 319 units) 111 27.5 (26.5)
12 Income mix was based on current total unit production, so one site without any current units cannot be included in this analysis. Also, an All But The Poorest site was excluded (N = 235). National Initiative on Mixed-Income Communities 43 | P a g e Case Western Reserve University
Age Table 25 illustrates the differences in return rates by age. As would be expected, older sites have a much higher return rate than newer sites.
Table 25: Comparison of return rates by age (N = 235)
Return Rate (%)13 n M (SD) Older sites (≥ 12 years) 121 33.8 (27.9)
Newer sites (≤ 11 114 21.1 (20.7) years)
Funding Sources In this section we describe sources of redevelopment budgets and the expended funds for projects.
Tables 26 and 27 provide the average, median and range for HOPE VI funds, Other Public Housing, Other Federal and All Non-Federal funds as described in quarterly project reports. Across developments the expended HOPE VI funding ranged from $14.6K to $67.7M with an average of $22.9M and a median of $20M. The expended total funding for HOPE VI projects ranged from $0.9M to $241.5M with an average of $65.3M and a median of $53.2M.
Table 26: Funding Sources across Developments (N = 260)
M Median Min if Min > 0 Max HOPE VI $22.9M $20M $14.6K $67.7M Other Public $6.6M $2.1M $6K $138.3M Housing Other Federal $6M $2.4M $13.8K $111.2M
All Non-federal $29.8M $20.8M $2.7K $174.1M
Total $65.3M $53.2M $0.9M $241.5M
To understand the trajectory of funding over the course of the HOPE VI program, Figure 11 below shows funding for HOPE VI projects by award year. The expended HOPE VI funding ranged from $66.2M to $566.1M with an average of $333M and a median of $446M. The
13 Among the 236 sites, one site did not provide its first unit availability date, which led it to be excluded from this analysis (N = 235). National Initiative on Mixed-Income Communities 44 | P a g e Case Western Reserve University expended total funding for the projects ranged from $197.1M to $1.7B with an average of $948.7M and a median of $1.1B, Funding was highest in the 2001 award year and dramatically decreased between 2003 and 2004 with the lowest funding in 2006 and 2009 and relatively low funding through 2010.
Figure 11: Expended HOPE VI funding by award year
$2,000,000,000 $1,800,000,000 $1,600,000,000 $1,400,000,000 $1,200,000,000 $1,000,000,000 $800,000,000 $600,000,000 $400,000,000 $200,000,000 $0 199319941995199619971998199920002001200220032004200520062007200820092010
Total Funding HOPE VI Funding
In addition, Figure 12 shows the different types of funding sources by award year. HOPE VI funding decreased between 1993 and 1996 award years, increased through 1999, decreased gradually through 2003 and then had a huge drop in 2004, with funding at a low level through 2014. Other public housing funding was highest for the 1996 award year and then remained relatively low through 2014. Other federal funding was consistently low from 1993 to 2014. The biggest changes were seen in all non-federal funding. There was a huge peak in the 1996 award year with a drop in 1997, followed by another big peak through 2001. This peak was followed by a drastic decrease in the 2004 award year, followed by another drop in 2006. There was an increase in the 2007 award year, but it was followed by another drop in 2009. The 2010 award year had an increase, which may have been a sign of recovery from the Great Recession that was responsible for the low non-federal funding from 2004 to 2007.
National Initiative on Mixed-Income Communities 45 | P a g e Case Western Reserve University
Figure 12: Expended funding sources by award year
$1,000,000,000 $900,000,000 $800,000,000 $700,000,000 $600,000,000 $500,000,000 $400,000,000 $300,000,000 $200,000,000 $100,000,000 $0 199319941995199619971998199920002001200220032004200520062007200820092010
HOPE VI Other Public Housing Other Federal All Non-federal
National Initiative on Mixed-Income Communities 46 | P a g e Case Western Reserve University
Comparative Analysis of Funding In this section we compare HOPE VI funding and total funding by region, income and tenure mix, size and age.
Region Table 27 illustrates the differences in HOPE VI funding and total funding by region. HOPE VI funding was similar by region and there were expected differences in total funding by region. Overall, the Northwest (Region 10) had larger amounts of total funding than any other regions.
Table 27: Comparison of average HOPE VI funding and total funding by region (N = 260)
HOPE VI Funding Total Funding
n M (SD) M (SD)
Northwest 12 $27M (11.3M) $105.5M (71.6M)
New England 15 $23.1M (12.7M) $79.9M (39.2M)
New York-New Jersey 24 $24.4M (14.9M) $78.6M (47.4M)
Rocky Mountain 5 $17.6M (10.5M) $76.1M (64.4M)
Pacific 18 $21.7M (14M) $67.7M (34.7M)
Total 260 $22.9M (11.3M) $65.3M (42.9M)
Midwest 41 $24.9M (11.1M) $63.4M (43.3M)
Southwest 19 $25.7M (11.8M) $62.5M (42.2M)
Southeast-Caribbean 75 $21.8M (91.7M) $58.1M (38.6M)
Mid-Atlantic 43 $21M (93.9M) $57.5M (33.4M)
Great Plains 8 $22.4M (18.3M) $52.6M (38M)
National Initiative on Mixed-Income Communities 47 | P a g e Case Western Reserve University
Income/Tenure Mix
HOPE VI and Total Funding by Income Mix We compared HOPE VI funding and total funding by income and tenure mix (Table 28). To compare HOPE VI funding and total funding by income we again used the Vale and Shamsuddin (2014) typology. HOPE VI funding was slightly higher in Broad Continuum sites, and total funding was highest in All But The Poorest sites followed by Broad Continuum sites. Notably, HOPE VI funding of the Broad Continuum sites varied much more widely compared to other income mix types. Also, the average total funding of the Narrow Low-Income sites was smaller than those on the Broad Continuum and the Polarized Bimodal sites.
Table 28: Comparison of average HOPE VI and total funding by income mix (N = 258)
HOPE VI Funding Total Funding n M (SD) M (SD) Narrow Low-Income 179 $22.5M (11.4M) $57M (36.5M)
Broad Continuum 66 $23.9M (94.8M) $86.9M (50.3M)
Polarized Bimodal 13 $27.7M (15.9M) $78.6M (43.8M)
Total 258 $23.1M (11.2M) $65.7M (42.8M)
HOPE VI and Total Funding by Tenure Mix We also compared HOPE VI funding and total funding by tenure mix. Sites without homeownership units had smaller HOPE VI and total funding compared to those with homeownership units. Table 29 below shows differences between HOPE VI funding and total funding by tenure mix.
Table 29: Comparison of average HOPE VI and total funding by tenure mix (N = 260)
HOPE VI Funding Total Funding n M (SD) M (SD) Rental-Only 76 $21.7M (13.6M) $54.7M (35.5M) Developments Mixed Tenure 184 $23.5M (10.2M) $69.7M (44.9M) Developments
Age On average, the older sites (n = 127) received $26.4M of HOPE VI funding and the newer sites (n = 131) received $19.8M of HOPE VI funding.
National Initiative on Mixed-Income Communities 48 | P a g e Case Western Reserve University
Table 30: Comparison of average HOPE VI and total funding by age (N = 235)
HOPE VI Funding Total Funding n M (SD) M (SD) Older sites (≥ 12 years) 127 $26.4M (13M) $66.2M (43.4M) Newer sites (≤ 11 131 $19.8M (81.2M) $65M (42.6M) years)
National Initiative on Mixed-Income Communities 49 | P a g e Case Western Reserve University
Community and Supportive Services (CSS) Thus far in this report we have described the unit production achieved by the HOPE VI program, fulfilling the bricks and mortar goals of improving the physical conditions of the housing developments. In addition to these goals, the HOPE VI program was committed to providing support for households as they transitioned from the high-poverty developments and faced new opportunities in redeveloped or alternative housing. In their 2000 book HOPE VI: Community Building Makes a Difference, Naparstek and colleagues describe how HOPE VI was designed to help residents move out of public housing to a better life through self-sufficiency programs such as literacy training, job preparation, training, and retention, personal management skills, daycare, youth activities, health services, community policing or security activities, and drug treatment.
According to the HUD report “Community and Supportive Services for Original Residents: General Guidance for HOPE VI Program” (2000), CSS program funds were for original residents of the public housing (regardless of whether they returned to the site after revitalization) as well as for households that later moved into the revitalized site. Household needs were assessed prior to redevelopment and services were intended to be flexible to each household. Services were also meant to help position residents to return to the redeveloped housing once it was complete. This often meant adhering to stricter requirements and screening criteria that had not been required in the former public housing context. For example, there was often much stricter attention to lease compliance and sometimes additional requirements like background checks, credit checks and drug testing. These requirements, in addition to many other individual, community and structural barriers, often meant families were not able to return to the development once it was redeveloped (Joseph and Chaskin, 2012; Popkin et al., 2004). Participation in services was voluntary. The flexibility with which grantees had to shape their programs resulted in significant differences in CSS from site to site, though each site was expected to follow these general stated principles: Services to help residents make progress toward self-sufficiency Services designed to meet individual family needs Linkage to relocation with informed choice Community building Management monitoring and evaluation
Below we describe key findings from our analysis of HOPE VI revitalization grant CSS program reports as of the fourth quarter of 2014 (N = 260). It should be noted that we refer to these programs using the language and category labels provided in the HUD reports. CSS program reports include information on the number of enrollments for: Employment preparation/placement/retention Job skills training programs High school or equivalent education Counseling programs
National Initiative on Mixed-Income Communities 50 | P a g e Case Western Reserve University
Transportation assistance Child care Substance abuse programs and English as a Second Language (ESL) courses
Reports include information on the number of successful program completions for job skills training programs and high school or equivalent education. In addition, the reports provide enrollment and completion information about three key areas of the program: employment, economic development and homeownership.
In our descriptive analysis we first provide a nationwide description of the CSS program describing the number of enrollments, successful completions of programs and caseloads. This national description includes information about pre-revitalization residents and new residents of occupied housing. This leads to an individual site analysis of caseloads for pre-revitalization and post-revitalization residents. Next we describe the enrollment and completions of services. We finish with a comparative analysis of CSS programs offered by HOPE VI sites.
National Initiative on Mixed-Income Communities 51 | P a g e Case Western Reserve University
Nationwide The CSS program reports provide information on program completions and enrollments. These reports identified completions and enrollments as of 2014 in addition to a goal established for the project. Table 31 summarizes key categories of the nationwide CSS information.
Table 31: Nationwide CSS information (N = 260)
2014-Q3 Goal Employment preparation/placement/retention 86,484 46,465 Transportation assistance 75,754 35,406
Counseling programs14 62,150 38,004 Job skills training programs 35,180 25,998 Number of Enrollments Child care 30,204 25,204
High school or equivalent education 18,470 13,616 Substance abuse programs 6,302 5,762
English as a Second Language (ESL) courses 4,232 3,801
Number of Successful Job skills training programs 18,610 15,474 Completions High school or equivalent education 5,322 6,827
Total new job placements 29,859 18,780 Employment Caseload currently employed 2,170 20,141
Caseload employed 6 months or more 1,397 15,055
Number enrolled – entrepreneurship training 3,752 3,517 Number completed – entrepreneurship training 1,652 1,999
Economic Development Resident employment in these businesses 777 1,310 Resident-owned businesses started 658 834
Resident-owned businesses – non-PHA funds 358 475
Number enrolled – homeownership counseling 16,707 13,908 Homeownership Number completed – homeownership counseling 7,229 7,682
Number purchasing a home 3,315 4,133
14 Counseling programs include programs designed to support and assist individuals dealing with personal or family problems, such as mental health issues, parenting skills, and family budgeting among others. Multiple enrollments across different counseling programs are counted here. Employment, substance abuse, and homeownership counseling are excluded in this category. National Initiative on Mixed-Income Communities 52 | P a g e Case Western Reserve University
Services with the highest number of enrollments were employment preparation/placement/ retention, transportation assistance, counseling programs, job skills training programs and childcare. High school or equivalent education, substance abuse programs and English as a second language courses had the lowest number of enrollments. These services all exceeded their enrollment goals. Job skills training programs exceed their completion goals while high school or equivalent education did not.
In the area of employment, despite exceeding the number of total new job placements, which includes all types of employment (full-time, part-time, seasonal and temporary), the caseloads currently employed and employed 6 months were far below their goals.15 While part of this shortfall might be due to a lack of accuracy caused by poor tracking and reporting, this could also be a clear indication of the difficulty of promoting job retention. In economic development enrollment, entrepreneurship training exceeded the enrollment goal but fell short on completion. There were also fewer resident-owned businesses or residents employed by those businesses than hoped. For homeownership counseling, enrollment exceeded the goal but the numbers of residents who completed counseling and purchased a home were less than expected.
In addition to providing information about key areas of the CSS program, the program reports provide information about the transition of residents from their former public housing to their current housing. Once sites were complete the program re-engaged with residents and began tracking their progress when they returned. Table 32 summarizes the nationwide caseload of pre- revitalization residents and new residents.
15 Caseloads currently employed: The number of individuals from the current caseload who are currently employed as of the last day of the reporting period. Caseloads employed 6 months (or more): The number of individuals from the currently employed (above) that have been employed 6 months or more as of the last day of the reporting period. The time between multiple jobs within a 6 month period can be no longer than 2 weeks to count as continuous employment. National Initiative on Mixed-Income Communities 53 | P a g e Case Western Reserve University
Table 32: Nationwide CSS caseload of pre-revitalization residents and new residents (N = 260)
2014-Q3 Original potential caseload 95,351 Residents ages 0-5 at the time of grant award 1,110 Services not accepted 11,108
Services not needed or authorized 14,563 Derivation of current Services no longer needed 14,069 caseload16 Moved out, unable to locate 14,854
Pre-Revitalization Permanent relocation, service handoff 4,481 Residents Cumulative additions Caseload 18,177 Current caseload 54,453 Still on-site, pre-revitalization 5,869 Relocated – other public housing 13,680 Location of current Relocated – Section 8 certificates/vouchers 18,008 caseload Relocated – not HUD-assisted 5,423
Returned after revitalization 11,473 New potential caseload 24,359 New Residents- Post- Residents re-occupied to the new site ages 0-5 After 766 revitalization Revitalization caseload Services not needed, accepted, or authorized 5,747 Caseload Current caseload 18,612 Total Current Caseload17 73,065
16 Derivation of current caseload: Derivation refers to how the caseloads were categorized. Original potential caseload: All individuals between the ages of 19-64 at the time of grant award Services not accepted: The total number of people who currently refuse services Services not needed or authorized: The total number of people who currently are not tracked because they are seriously disabled and unemployable, turned 65 years of age, were evicted or are deceased Services no longer needed: The total number of people who currently are not tracked because case managers have determined that they no longer need services Permanent relocation, service handoff: The total number of people who currently are not tracked because they made a permanent move and the HOPE VI staff arranged services with other providers in their new location Cumulative additions: Individuals who because eligible for CSS by (a) joining a CSS eligible household during the life of the grant, (b) moving into the site prior to relocation or (c) turning 19 years old Current caseload: A sub-total of the total caseload that includes all active original residents and cumulative additions. 17 Total Current Caseload: It includes all currently active original residents, cumulative additions and new residents living in revitalized units. National Initiative on Mixed-Income Communities 54 | P a g e Case Western Reserve University
Of the 95,351 residents in the original pre-revitalization potential caseload, 11.6% did not accept services and 15.3% did not need or were not authorized for services. By 2014, 14.8 % were categorized as no long needing services. The total national HOPE VI CSS caseload of about 73,000 in 2014 included 25% new residents to the developments and 75% original residents.
CSS Services across Sites The following table indicates which sites had each of the eight CSS services.
Table 33: CSS Services
n % Employment 248 95.4 preparation/placement/retention Job skills training programs 247 95
High school or equivalent education 246 94.6
Counseling programs 240 92.3 CSS Service Transportation assistance 238 91.5
Child care 237 91.2
Substance abuse programs 211 81.2
English as a Second Language (ESL) courses 110 42.3
Note: 260 total HOPE VI sites
National Initiative on Mixed-Income Communities 55 | P a g e Case Western Reserve University
IV. Discussion This report provides a detailed descriptive summary of the production and financing of HOPE VI projects nationwide from 1993 through 2014. We addressed the overarching research question: What is the income and tenure mix of housing units that have been produced through the HOPE VI program? In addition, we answered questions about the nature of HOPE VI re-occupancy, timeframes of production stages, financing and the Community and Supportive Services (CSS) available to residents of HOPE VI developments. We now review our major findings before concluding with implications for research and policy moving forward. Our findings confirm many longstanding concerns about the program such as the reduction in the overall public housing stock and the low return rates of original residents, while uncovering some compelling insights about the program such as the overall focus on low-income housing production and the extent of the production of public housing homeownership units.
Production Our analysis of the HOPE VI program production provided an overview of projected and actual unit production including a comparison by region, age, size and income and tenure mix, an analysis of re-occupancy by former public housing residents, and an analysis of production timeframes. Overall, the main HOPE VI redevelopment product was replacement public housing units, complemented primarily with the inclusion of affordable housing units and, in lower proportions, with market-rate units. Almost half of the redevelopments did not include any market-rate housing. The vast majority of production was rental units. We found it notable that a substantial number of homeownership units designated for public housing residents have been produced across almost half of the sites. Sites with only rental units tended to be smaller than mixed-tenure sites, and included fewer affordable and market-rate units.
Using Vale and Shamsuddin’s 2014 income mix typology, we found that well over two-thirds of the sites have a Narrow Low-Income mix of public and affordable units and only about a quarter have a Broad Continuum mix of public housing, affordable and market-rate units. The program is widely criticized for the major reductions in the number of public housing units and our analysis documents this drastic reduction. Almost the same number of units were constructed as the number of public housing units that were demolished. However, due to the incorporation of affordable and market-rate units, 43,274 units have been lost from the public housing stock. Our analysis also shows that the majority of HOPE VI projects have focused primarily on producing housing for low-income households, replacing over half of the public housing units and then building out another third of the developments with subsidized “affordable” units.
Looking at production over time, older sites produced more public housing units than newer sites and there was a greater focus on producing a more diverse income mix at developments as the program progressed over the years. The greatest demolition and production of units occurred in the first ten years of the HOPE VI program. Starting in 2003, there was a leveling off of demolition, relocation and construction and a sharp drop in production of homeownership units.
National Initiative on Mixed-Income Communities 56 | P a g e Case Western Reserve University
This reflects the fact that during President George W. Bush’s administration, the HOPE VI program was threatened with elimination each fiscal year. Although Congress successfully continued funding the program, the funding was significantly cut during this time and continued to be reduced through the final award year in 2011.
The data we received included only the most recent agreed-upon projections between the grantees and HUD, thus we were not able to analyze original production projections against actual production. In most cases, grantees have met their most recent projected production goals. As would be expected given the timing of the Great Recession and associated housing market crash, where production goals were not met, it was homeownership units that most often fell short, particularly market-rate homeownership. Thus the intended mix of incomes was not achieved at many sites. This can be read at least two ways. For those who feared that HOPE VI mixed-income projects would create environments in which public housing residents in particular, and low-income household in general, were a significant minority, this has not come to pass. On the other, to the extent that the success of the mixed-income strategy is considered to depend on a broader mix of incomes and a critical mass of higher-income residents, including homeowners, then this was not widely achieved through the HOPE VI program.
Another major criticism of the program is that original residents were not the ultimate beneficiaries of the redevelopment and return rates were extremely low. Among the total units redeveloped through HOPE VI, only about a fifth have been occupied by original residents of the development. And of the original residents relocated from the old developments, on average only a little over a quarter have returned to the new developments. Almost a fifth of residents were evicted, were deceased, or otherwise left the public housing development during redevelopment. Returns were similar regardless of development unit size, but newer sites have had lower return rates. The Southeast region had the lowest return rates despite the fact that it was awarded the most grants.
The number of phases and the time it took for each phase of development to begin and end had a major impact on how the HOPE VI program affected residents. The long periods between relocation, construction and occupancy meant an extended waiting period for those residents who hoped to move back to the new developments and likely contributed to the low return rates. We found that the relocation phases took the longest average amount of time (694 days) followed by construction (667 days), demolition (516 days) and occupancy (260 days).
Financing HOPE VI funds were used to leverage significant amounts of other public and private funds. The $6B of expended HOPE VI funds leveraged $11B non-federal funds to complete funding for sites with a total of $17B expended from all sources. For every dollar of HOPE VI funds, about $1.8 dollars were leveraged. The average total funding per site was $65.3M. The Northwest region achieved the highest leverage ratio of 1:3.3. Narrow Low-Income developments, which
National Initiative on Mixed-Income Communities 57 | P a g e Case Western Reserve University have only public housing and affordable units, generally had less funding and were not as able to leverage additional funds as much as projects that included market-rate or homeownership.
Community and Supportive Services The Community and Supportive Services program provided services to original residents of the public housing developments as well as families that later moved into the revitalized site. We analyzed caseload output data for the program, which included goals and outputs for each service. Only 31.8% of pre-revitalization residents received services. Services with the highest number of enrollments were employment preparation/ placement/retention, transportation assistance, counseling programs, job skills training programs and childcare. Overall, CSS programs at sites met or exceeded most enrollment output goals in these areas. Generally sites fell short of completion goals for high school or equivalent education. In the area of employment, despite exceeding the enrollment goals for total new job placements, the number of residents currently employed and the number employed for six months were far below their goals. In the economic development service area, enrollment goals for entrepreneurship training were exceeded, but the total number of residents who completed programs in this area was less than the goal. There were also fewer resident-owned businesses or residents employed by those businesses than hoped. Residents successfully enrolled in homeownership counseling, but the number of residents who completed counseling and the number who purchased a home were less than expected.
Study Limitations There were some important limitations to this study. The findings are drawn from HOPE VI administrative project reports from 1993 to 2014 and include quantitative data on production, financing and the CSS program and thus provide a numerical documentation of the program. But given the limited information available in the report, we were constrained to providing basic descriptive and comparative statistical analyses of these data. There was a text field for qualitative comments in the CSS reports, but those data were of limited depth, consistency and utility. Also as stated earlier, although these reports provide “projected” estimates regarding financing and unit production these estimates were updated over the course of the program and thus do not provide an accurate beginning time point for comparison.
National Initiative on Mixed-Income Communities 58 | P a g e Case Western Reserve University
V. Conclusion and Implications for Research and Policy The purpose of this research was to understand the income and tenure mix of housing units that have been produced through the HOPE VI program. We also explored the pace and duration of the various development phases, the extent of relocation and return of original residents, the funding leveraged for the program, and the services provided through the CSS component of the program. We learned that HOPE VI production has been predominantly public housing and subsidized rental housing, with less production of market-rate rental and homeownership units than intended. Although it produced higher-quality replacement public housing in more mixed- income environments, the HOPE VI program also substantially decreased the stock of public housing units. Furthermore, extended construction and occupancy timeframes and stringent re- occupancy requirements may have prevented many residents from successfully returning to the revitalized housing. HOPE VI funds successfully leveraged substantial levels of private funds for development. But due to the decrease in federal funding over time, unit production levels were not sustained over the life of the program. CSS programs generally achieved many of the stated enrollment goals but were limited in the number of residents they could serve and fell short of the stated goals in several important areas such as high school equivalency, sustained employment, and entrepreneurship and homeownership training. These findings suggest a number of implications for policy and further research.
Implications for Policy Stepping back from this largely descriptive analysis, we can draw from these findings to suggest some key areas for continued policy focus and improvement.
Balancing dual priorities: ending segregation and concentrated poverty and increasing affordable housing for the poor. This is an enduring strategic tension in this arena of housing policy. There are conflicting, legitimate policy imperatives. It would be ideal to be able to produce more high quality public housing and facilitate access to more vibrant, socioeconomically diverse neighborhoods. No housing policy has yet resolved this dual challenge and the major policy approaches: mixed-income redevelopment, housing choice vouchers, and the Rental Assistance Demonstration program each have operational advantages and important downsides. Policymakers must continue to be as intentional and comprehensive as possible in their efforts to use existing public resources to maximize the provision of housing for the poor while leveraging private sector resources to generate investments in developments and their surrounding neighborhoods. Clarity about the intended balance of these conflicting goals and vigilant accountability will be key.
● Managing the market risks of privatization. The mixed-finance approach to HOPE VI and the intentions of including market-rate rental and homeownership made the redevelopment efforts extremely vulnerable to market conditions. The result, evident in
National Initiative on Mixed-Income Communities 59 | P a g e Case Western Reserve University
the early phases of the program and then exacerbated by the housing market crash, was extended delays in unit production and ultimately substantial shortfalls in the production of market-rate units. Policymakers should consider to what extent the market-rate shortfalls affected program “success” in various local contexts and examine ways to offset the market risk; for example, might there be a benefit to smaller redevelopment phases and increased capacity-building and engagement of non-profit entities and community-based organizations?
● Increasing return rates. Clearly a major program shortcoming was the limited proportion of original residents who benefited from the new, higher quality housing and living environments. While some proportion of the non-returners possibly used the relocation opportunity for upward mobility to a low-poverty neighborhood, the literature suggests mixed results for those who did not return and for voucher holders in general. The policy implications include increased resources and attention to the relocation support process and strategies to make return easier such as smaller redevelopment phases and phased relocation onsite or in close proximity.
● Program enrollment is not sufficient: Providing the requisite support services for positive outcomes. The CSS data analyzed here provide a severely limited purview into the details of service provision and results. However it is clear from the program reports that while the HOPE VI grantees were able to exceed their enrollment goals in many cases – indicating that they had achieved the objectives of linking with local partners and programs that were offering the forms of support needed – it appears evident that those enrollments generally did not turn into sustained engagement nor meaningful outcomes for participants. What is the scope, quality and duration of support needed to help residents affected by a mixed-income redevelopment move toward self-sufficiency?
Implications for Further Research Given the continued national and local investment in the mixed-income redevelopment approach with now over 100 projects with Choice Neighborhoods Initiative implementation and planning funding and an array of other local mixed-income efforts, the scope and duration of the HOPE VI program makes it an important enduring learning resource for the housing and social policy fields. There are a number of research topics that could be pursued, either from further data collection and analysis of HOPE VI projects or from research on Choice Neighborhood grantees and other emerging mixed-income redevelopments.
● To fully understand the projected versus actual unit production through the HOPE VI program – and thus inform future development efforts and negotiations – an analysis using additional information about original projected goals is warranted.
National Initiative on Mixed-Income Communities 60 | P a g e Case Western Reserve University
● The analysis here was technically about the mix of subsidy types – public housing, “affordable,” and market-rate but not about actual income levels. A deeper understanding of the income levels of residents in each band of subsidy would provide far better information about what households are benefiting from the housing and what the income mix continuum looks like in various contexts.
● Stronger performance measurement and reporting of resident outcomes would provide a critical knowledge base on a number of questions including: ● The results of various levels and combinations of supportive services. ● How the outcomes for residents compare across different levels of income mix, for example Narrow Low-Income developments compared with Broad Continuum developments. ● How outcomes for original residents who return compare with those who don’t. ● How outcomes compare among original residents who return, newcomers to the public housing replacement units, residents of the affordable units, and residents of the market-rate units. ● How resident outcomes vary with factors such as the size of the development and the tenure mix at the development.
The nature, outcomes and impact of public housing homeownership programs is an under-examined area, including which residents qualified and participated and their outcomes.
There were a range of relevant issues raised in the growing literature on mixed-income redevelopment that were well beyond the scope of the basic data available for this study but that should be considered in future research in this area including, for example: physical design, layout and integration, property management, governance and participation, and community building and social cohesion.
The focus here, given the data available, was on the developments and not the broader neighborhood context. But the surrounding neighborhoods are an important dimension to be included in future research, both as a context that shapes the strategies and outcomes in particular developments as well as a unit of change impacted by the redevelopment.
Now that this comprehensive HOPE VI dataset has been extracted and compiled it should serve as a resource to others looking to investigate unit production and other dimensions of the HOPE VI program. We plan to make it available for further analysis as part of the online Mixed-Income Development Database hosted by the National Initiative on Mixed-Income Communities website at nimc.case.edu.
National Initiative on Mixed-Income Communities 61 | P a g e Case Western Reserve University
Acknowledgments We thank our colleagues Dr. Claudia Coulton and Dr. David Crampton for their advisory roles on this research study. We thank research assistants Biwen Liu, Siyang Sun and Michael Salwiesz for their data compilation work. We thank our partners at the Resilient Cities and Housing Initiative at the Massachusetts Institute of Technology led by Dr. Lawrence Vale for their collaboration and insights in extracting, compiling and analyzing this dataset. We thank HUD for the Research Partnership grant that made this study possible and in particular Ron Ashford and Lars Gnissen for their help with our orientation to the dataset and Madlyn Wohlman-Rodriguez for her encouragement and responsiveness in supporting this research. We also thank Madlyn and the HUD staff who provided detailed feedback on an earlier draft.
National Initiative on Mixed-Income Communities 62 | P a g e Case Western Reserve University
References
Brophy, P. C., & Smith, R. N. (1997). Mixed-income housing: Factors for success. Cityscape, 3(2), 3-31. Chaskin, R. J., & Joseph, M. L. (2015). Integrating the inner city: The promise and perils of mixed-income public housing transformation. Chicago, IL: University of Chicago Press. Cisneros, H. G., & Engdahl, L. (Eds.). (2009). From despair to hope: HOPE VI and the new promise of public housing in America's cities. New York, NY: Brookings Institution Press. Fosburg, L., Popkin, S. J., & Locke, G. (1996). An historical and baseline assessment of HOPE VI, Volume I. Washington, DC: U. S. Department of Housing and Urban Development. Holin, M. J., Buron, L., Locke, G., & Cortes, A. (2003). Interim assessment of the HOPE VI program: Cross-site report. Washington, DC: U. S. Department of Housing and Urban Development. Joseph, M. L. (2006). Is mixed‐income development an antidote to urban poverty? Housing Policy Debate, 17(2), 209-234. Joseph, M. L. (2013). “Cityscape” Mixed-income symposium summary and response: Implications for antipoverty policy. Cityscape, 15(2), 215-221. Joseph, M. L. & Chaskin, R. J. (2012). Mixed-income developments and low rates of return: Insights from relocated public housing residents in Chicago. Housing Policy Debate. 22(3), 377-406. Joseph, M. L., Chaskin, R. J., & Webber, H. S. (2007). The theoretical basis for addressing poverty through mixed-income development. Urban Affairs Review, 42(3), 369-409. Khadduri, J. (2001). Deconcentration: What do we mean? What do we want? Cityscape, 5(2), 69-84. Kleit, R. G. (2005). HOPE VI new communities: Neighborhood relationships in mixed-income housing. Environment and Planning A, 37(8), 1413-1441. Naparstek, A. J., Freis, S. R., Kingsley, T. G., Dooley, D., & Lewis, H. E. (2000). HOPE VI: Community building makes a difference. Washington, DC: U. S. Department of Housing and Urban Development. Popkin, S. J. (2010). A glass half empty? New evidence from the HOPE VI Panel Study. Housing Policy Debate, 20(1), 43-63. Popkin, S. J., Katz, B., Cunningham, M. K., Brown, K. D., Gustafson, J., & Turner, M. A. (2004). A decade of HOPE VI: Research findings and policy challenges. Washington, DC: Urban Institute. Popkin, S. J., Theodos, B., Getsinger, L., & Parilla, J. (2010). An overview of the Chicago family case management demonstration. Washington, DC: Urban Institute. Turbov, M., & Piper, V. (2005). HOPE VI and mixed-finance redevelopments: A catalyst for neighborhood renewal. Washington, DC: Brookings Institution.
National Initiative on Mixed-Income Communities 63 | P a g e Case Western Reserve University
U.S. Department of Housing and Urban Development. (2000). Community and supportive services for original residents: General guidance for HOPE VI program. Retrieved from http://portal.hud.gov/hudportal/documents/huddoc?id=DOC_9840.pdf U.S. Department of Housing and Urban Development. (2003). Guidance for PHAs developing a Section 32 homeownership plan. Retrieved from http://www.hud.gov/offices/pih/centers/sac/homeownership/section32deskguide.pdf U.S. Government Accountability Office. (1997). Public housing: Status of the HOPE VI demonstration program (RCED Publication No. 97-44). Gaithersburg, MD: U.S. General Accounting Office. Vale, L., & Shamsuddin, S. (2014). All mixed up: Defining mixed-income in public housing redevelopment (Working Paper). Massachusetts Institute of Technology.
National Initiative on Mixed-Income Communities 64 | P a g e Case Western Reserve University